Genomic selection, which is currently being implemented within several species, is often considered too expensive for aquaculture, due to the costs of genotyping many individuals for high density SNP marker panels. With rapidly dropping genotyping cost th e prospect of expansive use of this new powerful technology is however looking promising. Within family genomic selection is a strategy where combined use of molecular and pedigree information reduce the required marker density. This can be implemented at low risk within the current breeding programs, as it only affects the within family component of the breeding value, while the between family component is estimated by conventional methods. However, proper implementation requires increased knowledge amon g the breeding companies of how to use these new genomic tools. At present, the effects of the number of animals genotyped and the SNP density on different traits are unknown. This project aims to optimise within family genomic selection for selected dise ase and quality traits which is an important first step towards implementation.
The most cost efficient SNP panel for each company will be identified through cross-validation of breeding values. Simultaneously, breeding value estimation methods and selec tion strategies will be evaluated and compared for each breeding company. This will result in optimised strategies on which fish to genotype for which markers and how to estimate breeding values to obtain maximum genetic gain. Possibilities for reduced co sts through cooperation between companies on the genotyping will be evaluated. Results from different companies will increase the understanding of how different population structures and breeding goals affect how genomic selection should be implemented. I mprovements on the scale indicated by simulations can add up to become very significant economic factors over a few breeding generations considering the great volumes of Atlantic salmon produced.